ASML Holding N.V. (ASML) Earnings Call Transcript & Summary

I'm Amit Harchandani, Head of Citi's European Tech Research team and your host for this virtual fireside chat session on ASML as we kick off Citi's 2021 Global Technology Conference this morning. Thanks for joining us, and I do hope you and your loved ones are safe and healthy. Before I move on to introducing our main speaker, I would like to highlight that we are keen to take questions from investors joining us on this fireside chat. So please do send those over to [email protected]. Alternatively, an even better manner would be to just put it in the chat window that you may see on your screen. So with that bit of housekeeping out of the way, it is my pleasure to introduce our main speaker for the keynote. ASML, CFO; Roger Dassen, accompanied as ever by Skip Miller and Pete Convertito from the ASML Investor Relations team. Gentlemen, on behalf of Citi and its clients, thank you for joining us and supporting our conference every year.

Thank you, Amit. Our pleasure. Good morning, good afternoon, everyone.

So Roger, in terms of the plan for the session, lots to get through, and we aim to structure it across 6 broad topics. In the first 20 minutes, I'm keen for us to discuss ESG near-term demand and EUV, then in the second half of the session, I'm hoping for us to talk about Deep UV financials and round off with some big picture questions. So to get things going, Roger, I'm, as I said, keen to briefly talk about sustainability, which is becoming a critical topic for the investment community. Firstly, can you remind us of what the key sustainability targets are for ASML? And how are you progressing against those targets?

Absolutely, Amit. And I think as people know, we try to bucket the ESG goals that we have into 5 buckets, as a matter of fact, 5 ESG buckets. So one is on people, one is on innovative ecosystem, one is a responsible supply chain, one is on reuse, in essence, and circular economy, and one is on a climate carbon footprint, energy use. And for each of those 5 buckets, we have very concrete targets, also concrete targets for the '25 timeframe that we're working towards. These are visible on the website, visible in the integrated work for. But just to give you a flavor of, for instance, what we're trying to do. On the -- on energy use, we followed the Scope 1, Scope 2, Scope 3 definition that, I guess, most investors will be very familiar with, and we have very specific goals for those. So for instance, on Scope 2, we're essentially done. So starting 2020 on Scope 2, we have 0 emission because all the electricity and assets that we use is from renewables. On Scope 1, we have a detailed plan to make sure that we're completely emission-free on Scope 1 by 2025. And then finally, on Scope 3, which, of course, is quite a bit more complicated because it extends to the entire ecosystem, including the use of our tools by our customers. That's why I primarily focused on reducing the energy consumption of our own tool, for instance, and there we have targets. One target is that the energy consumption per wafer pass would be 60% lower by 2025 than what we would have in 2018. So for all of those, we have very concrete targets and we're measuring progress against those targets internally periodically and in the annual review, obviously, once every year.

Thanks for that overview, Roger. And maybe very briefly, if I could also comment upon how the core compensation of the top and the senior management is increasingly being linked to the achievability of the sustainability and inclusion targets?

Yes. So at this stage, 10% of the LTI is based on sustainability goal. And there, our achievement there, our performance there is really measured in terms of how we perform on the Dow Jones Sustainability Index. So how we progress against the absolute leaders on that front. So our relative performance to the leaders determines 10% of the LTI. That's the way it's going to impact in the compensation structure.

Noted. Noted. So let's maybe now move on, get into the thick of things, maybe getting into the near-term demand dynamics. For 2021 sales, your key assumptions are Logic up 35% year-on-year, Memory up 60%, and installed base management up 15% year-on-year. Are you as confident today around each of these targets as you were back in July? Or have things gotten even better?

I would say I'm as confident as we were in July. Also recognizing, Amit, I think we made that point also in July. The demand is there, no worries. It really depends on the ability to execute. So if anything, for this year, we are clearly supply constrained. So to be more confident than we were in July would be a bit of a stretch because we're executing, of course, against plan. But I think the numbers that we articulated in July are really what we think we can realistically do on the supply side. No less confident either. First, because we are executing according to plan. And secondly, because there is no sign whatsoever that on the demand side, that would be an issue. In fact, demand clearly outweighs for this year the supplier. So no, I would say as confident as we were in July that we're able to meet those objectives.

All right. Digging a bit deeper into your demand drivers, starting with maybe Memory. There is some investor concern about a down cycle in the second half of this year, potentially 2022. The jury is still out on that one. What is your own impression? And one of the points that comes up with, for example, could the availability of clean room for DRAM post a bump down the road for you in 2022?

Yes. So in general, I think, Amit, all the drivers that we saw for the Memory -- for the continued strength in Memory for us to remain intact. And when I talk about strength, first and foremost, I talked about the strength in DRAM because that's, I think, the lion's share of the demand that we'll have. We'll make a short comment on the later on. But if we look at DRAM, I think it's fair to say that the bit growth in the course of this year went from mid-teens to around 20 or even low 20s in the course of this year. So I think from that vantage point on DRAM, you see sustained demand and increase in bit growth. So I think that's an important driver. I think you also see that with the continued strength in Logic, you see that, that strength in Logic also sustainably drives the demand for Memory associated with that. So those 2 seem to go hand in hand. So systemically, we think that, that is moving well. There is a number of one-off effects for us. One is on the technology front, right, where we see the insertion of EUV into DRAM are getting stronger and stronger. We've covered that and we've looked at the -- we've talked previously about the comments that our customers make in that regard. So I think that is a strong addition for us. So irrespective, I would say, to a certain extent of the global demand for DRAM, you see that technology transition happening. And then lastly, I think it's also fair to see that in the China demand that we have this year. Although both fueled by growth in Logic and Memory, I would say that the China demand is particularly strong and the growth is particularly stronger on the Memory side than it is on the Logic side. And that would contain both DRAM and NAND. So I think from that vantage point, we see continued strength in the demand for Memory on our side. Specifically to your question on clean room, I hear those comments once in a while. To be honest with you, if you look at the developments of our customers, if you look at the clean room that is being built. And we both know some phenomenal projects that certain of our customers are engaging in and are in the process of completion. I don't think that, that is a very, very valid argument to assume that the memory demand will go down. I see still a lot of availability of clean room also in the memory space. So I wouldn't see that as a particularly confining factor.

Thank you, Roger. Good to get that one out of the way. Moving on, I guess, towards Logic now. The picture is getting a bit interesting with Intel looking to use TSMC for its graphic chip products in '22 and some non-CPU components in '23. Samsung obviously has its own ambitions. Help us understand the dynamics in the near term across your Logic customer base. How do you see the visibility? Where do you see greater traction? Any thoughts on Logic demand near term?

Yes. I mean, I would say, in general, all the comments that we've made in the past couple of months on Logic still apply, right? So the demand is clearly there, and it's fueled by the 3 trends that we've talked about in the past, which are particularly visible, I would say, on the Logic front. So on the one hand, there is some catch-up going on. There's some catch-up activity going on. This might be most particular in the -- on the more mature nodes with all the comments that customers, but also the customers of our customers and the end markets make on the chip shortage being sustained well into next year. We've all seen those comments in the past couple of weeks. So that trend does not seem to come to an end very, very soon. So that's one dimension. Much more important, I think, is the secular trend. That is also much more sustained. The secular trend behind the demand for Logic, which is everything associated with 5G, everything associated with high-performance compute. That trend is clearly -- and artificial intelligence, those trends are clearly there, and they seem to continue quite well. And then thirdly, of course, we have the -- what we call the technological sovereignty, where we see fabs being built in different regions, much more so than in the past. All the activities considered in the United States and in Europe, for instance, come to demand. So I think those are the 3 drivers. Specifically to your question, there's clearly competition going on, and there isn't a single day that somewhere you pick up an article that talks about the ambitions of one of the players be it the Taiwanese player, be it the Korean player or be it the player from North America in terms of leadership in the foundry business. And frankly, Amit, we love it. Competition improves the breed, but competition and really competing for the top position in foundry is, of course, something that we as an equipment manufacturer really like. Specifically to your question, some of the things that you see, of course, they are linked to, on the one hand, wanting to maintain or even increase market leadership of one of the players. And on the other hand, specific characteristics of the development and then ramp plan that some of the customers have. So that's why sometimes you can see that even though we are looking at fierce competitors, sometimes they just need each other and they outsource certain stuff to each other. I think that in and by itself is not new. But the landscape, I think, is really characterized by 3 formidable players with a lot of ambition and a lot of innovation power. And I think that's healthy for the ecosystem and ultimately also very healthy for us as an equipment manufacturer.

Noted, Roger. So listening to you, if I try to put all these blocks together, and I've seen a few investor questions come in that I'm trying to weave in here. How does the picture for 2020 to look overall to you? If we go a little bit further, are you in a position to give us a steer in terms of what the kind of growth level one would look at going into 2022?

So I think, in general, Amit, I think it's fair to say that we think the growth drivers for this year and for next year are strong. Also in light of the fact, as I mentioned, that demand this year probably outweighs what we can build. That gives us a good head start, I would say, for the year. But in general, I would say the demand is quite strong. You know that we've indicated that we're looking at capacity for EUV for next year of 55 tools. So 55 tools at over 160 ASP because that's typically what we indicate as the lower threshold, if you like, for the ASP for EUV. If you take that, and that gives you already about EUR 3 billion incremental revenue for next year for EUV. EUV this year expected to be around -- to land around EUR 6 billion. So if you do the math that I just give you, it gets you to around EUR 9-ish billion for EUV for next year. We think the demand for next year in DPV is going to be strong. But I should warn that this year, we tap into any reserve that we had at the company in order to be able to ship tools to the customers. And of course, we need to rebuild to a certain extent those reserves again next year. So even though we said that we're looking at increasing capacity also for DPV in the course of this year and also a little bit of next, that doesn't necessarily translate into an explosion in DPV sales because there is all this replenishment that still needs to happen of the reserves that we tapped into. But nonetheless, installed base typically will continue to grow at a pace that I think most people understand. You see a EUR 3 billion approximately uptick in EUV. I think that gives you a healthy outlook for the year without expecting miracles on the DPV side.

Got it. And just before we leave the near-term picture and dive a bit deeper into EUV, a quick clarification question that's come in, Roger. Could you quantify what's the level of reserves you have tapped into on the DUV side this year? Is there a number you could give around that?

No. I can't. We don't really want to give a number there. But it's been substantial. I mean, particularly on the optical side, we've really stretched it, really took -- it's particularly there that we took certain lenses that are typically necessary also for other purposes and really put them into manufacturing. So -- but it's been fairly substantial. I would also say, Amit, that this year looks to be a very good year in terms of also the mix within DPV, right, with quite a bit of immersion in there. And of course, for next year, we know there is also quite a bit of demand for drive. So all those dynamics make for a good year without a doubt for DPV next year. But again, I wouldn't expect complete miracles to happen that all of a sudden, the DPV business next year would explode in comparison to this year for those 2 reasons combined.

Got it. And suffice to say, you don't seem to be worried about an overcapacity scenario just yet at this point of time given the additional situation...

No, we're not worried about that because we do see demand being quite sustained. That's what we hear from customers. And also, Amit, I said it before, the incremental capacity that we built, the payback time for that is so short that to me, this feels like a no regret bet.

Perfect. Right. Diving deeper, EUV. You made a brief comment on 2022 earlier. You talked about the EUR 3 billion uplift in revenues. Your backlog at the end of Q2, if our calculations are not wrong, covered 80% of your shipment capacity for next year. Well, we are almost 2.5 months and into Q3. Would you be able to confirm if you're fully covered already for 2022 now?

Well, I think you got it right. So we have an EUR 11 billion backlog, EUR 3.7 billion of that we need for this year, so EUR 7.3 million remaining for even next year and beyond. It's also a little bit beyond indeed. So that means that you're not too far off with the 80%. I think that will fill up rather quickly. So I would say by the end of this -- order intake is lumpy for us. And I think sometimes that can be pretty spectacular as it was certainly last quarter. Sometimes it can be a little bit less and will be a little bit less, but we cannot repeat what we did last quarter time and time again. But I'm pretty confident that by the end of this year, we will not just be covered for the full year 2022, but that will also already have been a very healthy order intake for the year after.

Fantastic. I guess moving on to the supply side of the equation, if you're getting covered so nicely in terms of demand coming in, help us understand how do you go from 40 to 55 capacity next year to ship EUV. Is that an underlying assumption that you are targeting for cycle time reduction?

Yes, definitely. So that's one of the ways to get there. So there are 3 ways that we're looking into this. So the easiest one and the cheapest one, CFOs love that, is cycle time reduction because that's positive, of course, on all fronts. And I think we've made quite some good progress on that. Of course, we had the introduction this year of the D model. Of course, then with the introduction of a new model, the progress we made over the year, there is -- it stalls a bit because a new model, of course, takes a little bit of learning curve. But making good progress on that front. So the 17 weeks number that I've indicated before, I think that's again in sight. So by the end of this year, I would expect that also, for the D, we're going to be at 17 weeks of cycle time, which is a big improvement because it wasn't so long ago that we were talking about 30 weeks, right? So I think good progress on that front has been made. So that's one element. The second element is within your existing clean room, putting in more people and putting in more equipment, and that's exactly what we're doing. This is true for EUV, but also true for DPV, by the way. And that's true for us, and it also needs to be made true, of course, for the supply chain. And the third element is to really build incremental clean room, right? I mean those are the 3 ways to get there. And the first one is easiest, cheapest and it can be done very, very quickly. The last one to build incremental clean room either at our own premises and/or at the premises of our main suppliers. Of course, that takes longer for that to take effect. I think it's fair to say that when we look at a little bit longer term, so when we talk about the over 60 that we've talked about for '23 and then maybe even more in years thereafter, of course, we would also, at some stage, have to look at that. But I would say in 2 weeks' time, we're going to have the Investor Day. And there, we will have a bit more in-depth discussion of how we look at capacity also beyond the '22, '23 timeframe.

Right. Yes, that makes it a bit tricky, doesn't it? But I guess you talked about 60-plus fine. But along with the 60-plus comes the improvement in performance, the step-up targeted with the e-tool. And then, of course, there's high NA on the horizon. You put it all together, that's a lot of leading-edge bit growth coming out. Isn't there a risk of overcapacity?

We base our plans and also our capacity build plans, obviously, in what we hear in the marketplace. And we talked about the strength in Logic. We've looked at -- and we've talked about the insertion of DRAM into -- or of EUV into DRAM manufacturing. And both, I think, make for very strong demand. And after that, all the plans that are out there to also build fabs and leading fabs in, if you like, new territories. And I think that probably explains why in the years that we're talking about here, let's say, the '23 and beyond timeframe, that at least in the medium term, we do think that there is quite healthy demand. Of course, high-NA is going to kick in. But the numbers that high-NA is going to kick in, let's say, the '25 timeframe, as you will appreciate, are fairly small. So all the way through 2025, I think the workhorse on the EUV layers will definitely be low-NA. You are quite right that we have a very strong productivity road map. If you look at the D tool and then also the progress that we envisage for the E tool, there is a very strong productivity improvement that we plan on those road maps. But if we try to piece together everything we hear from customers, everything we see in terms of end demand, et cetera, and again, more on that in 2 weeks' time. But all of that combined leads us to believe that when we talk about over 60 for '23 and beyond that, there is a compelling case for those investments to be made.

Noted, Roger. And maybe just to round off the discussion on EUV very quickly. Could you give us a sense for where the installed base is today of EUV? And how are your customers thinking about options and upgrades for their installed EUV systems?

Yes. So this quarter, we passed the 100 tools. So that's -- so more than 100 EUV tools have now been installed. For different types, you have different solutions in terms of upgrades. I would say, broadly speaking, there are 2 upgrades that customers are interested in. One is to replace the existing, the older vessels with the modular vessel because we've talked about the modular vessel before. That has a big impact on the availability of the tool because you have the in-line refill. And it's far easier to swap out the vessel to clean, et cetera, et cetera, to clean the collector, et cetera. So it's a big improvement there with a number of percentage points increase in availability of tool, frankly as a result of that. So that's appealing for our customers and is being done. And then the second upgrade is on productivity, right? So just increasing with -- particularly with the product enhancement packages, as we call them, increase the throughput per hour for the tools. So those are the tools that are in high demand. Customers are looking at this, how they can schedule these planned downs for those machines in terms of maintenance and making them available for upgrades. But the interest is high. It's just that they need to look at one is it that they can do it because all of these machines are also running at a very, very high capacity. So that's why the -- and the utilization, so that's what they're really looking at what's the best plan for this to be done. But the interest in both is very, very high.

All right. Definitely look forward to hearing more in a couple of weeks on EUV. But let's now transition, as we are past the halfway stage, to the topic that's positively surprised us all this year, DUV. Based on your comments so far this year, it appears that the longer-term DUV contribution could be in the ballpark of EUR 10 billion or even higher versus previous expectations, which were seen sort of trending more towards EUR 5 billion, if I go back to the 2018 Capital Markets Day. Could you remind us of what are the key factors in your view that's changed? And what's the kind of level for DUV you might think going forward?

Yes. That is a big number, by the way. I think if you look at this year and you do the math, then you would see that we're looking for the combination of DUV and apps that we're looking at EUR 8.6 billion. If you just do the math in terms of the growth of EUV and the installed base that we get to EUR 8.6 billion. So EUR 10 billion is a big number. Nonetheless, also EUR 8.6 billion is a big number, right? And considerably higher to your point, then the EUR 5 billion that we talked about back then. I think just going back to a comment I made a couple of minutes ago, Amit, on the key drivers. The shortage that we're talking about, the very strong secular trend on the end markets. And I mentioned the 3 big ones there in particular. And frankly, also the tax severance. I mean, they play for EUV, but also play for DUV. So while on the one hand, the migration of certain critical layers from immersion to EV happened as planned, and you might argue even more so. It is clear that the demand, on the other hand, has become so strong for incremental capacity. And that has really outweighed the cannibalization effect, if you like, of immersion by EUV and then subsequently thereafter. So I think it's the same 3 trends that we talked about that also made for such a better picture in -- as we look at it today in comparison to where we are at the end of '18.

Noted. And then just to clarify, the real sort of sovereignty demand, if I could call it, really kicks in from 2023 onwards for you, right, looking at some of the [ plans ] out there.

I think that is realistic. If you look at the building plans that are out there, I think the lion's share of that you should see kick in into '23 timeframe.

Got it. And how does this change your strategic thinking around the supply chain, Roger, the DUV supply chain, are you planning R&D differently? Incentivizing the supply chain differently? How are you looking at it now?

So as it relates to R&D and manufacturing, a long-held view within this company that it's important to keep those 2 together because we firmly believe that the cycles of learning that you get from the practice in manufacturing and then building up modules. Back to the R&D department that, that cycles of learning should be short and should also be physically short. So that's the reason why it's been a long-held view of this company that those should be kept together. So for that reason, I wouldn't expect dramatic changes to be done. And also, we truly believe that IP and development should be kept as closely together as you can. So we're not suggesting to have IP and R&D competencies being widespread throughout the globe. We try to have very significant chunks of discrete R&D activity really put together and being very, very close to the place where the manufacturing happens. In terms of supply chain, it's important that we continue to align the supply chain and that -- but that can be done in many different ways. You've seen in the past that once in a while, we do an acquisition to the extent that we believe that, that is necessary. But I would suggest that, that really is the outlier, is an outlier strategy. In general, we work very, very hard with the main suppliers to make sure that their interests and our interests are as aligned as possible and that the required investments that are necessary to meet both the value proposition that we want to meet and the volumes that we're driving, that it's worth that is worthwhile both for the suppliers and for ourselves. And frankly, I think also, if I look at the ramp-up that we've been able to do this year and also the plans that we have for the future, I see that our supply chain is very, very much attuned to that joint interest.

Noted. The last one, to round off the DUV discussion. China obviously remains a critical driver. Could you give us a sense for what is the contribution this year? And how does China factor into your sort of DUV projections beyond 2021?

Yes. So China is clearly important. If you look at the system sales for the past 2 quarters of this year, you would see that it's somewhere 15%, 17% of system sales comes from China. And I think there is no reason to assume that, that will be dramatically different for the balance of the year. And as I mentioned, it's a good combination this year of both Memory and Logic. Whilst last year, it was primarily Logic. That was very, very strong. This year, there is a better balance where there's also quite some Memory business in there. And Memory, not just DRAM, but also NAND. So a big chunk of the NAND business that we do today really comes from China this year.

Roger, could you get -- that 15% to 17% you referred earlier going to China is all domestic China?

It's not all domestic China, but it's nearly all domestic China. Yes, this year, the lion's share of the sales into China is to domestic players. That is correct, Amit. So that's the picture, and that's a picture for this year. I would say for next year and for the years thereafter, based on all the conversations that we have, the number of new fabs that are being built in China, but also the demand forecast that we get from the large players, there's no reason to assume that, that growth curve is going to come to a grinding halt. Of course, we're all looking at the discussions between several countries. We're looking at the trade wars and all the consequences there. Of course, that's something that we have a very close eye on, what measures can be assumed there. But assuming that there is no draconian impact coming from that, I believe that looking at China and for it to bring something within this share between 15% and 20% for the years to come is a good asset.

Thank you, Roger. Moving on further to now talk about the financials with the CFO. Given the multiple tailwinds you've talked about into next year, you talked about the EUR 3 billion extra EUV systems revenue, the EUV service margins would improve, the mix would shift to 3600D, the NXT:2050 on the immersion side. What stops your gross margins from moving from 51%, 50% to this year to maybe 53% to 54% or even more in 2022?

Yes. I mean I'm not going to make any quantitative comments on that. Of course, again, on the Capital Markets Day, we're going to give some insight into how we look at gross margins beyond this year. But I think it is reasonable to assume that some of the drivers that you just hinted at would give us some positive tailwind indeed. So I think you nailed the key elements. I think the D tool and all of the EUV tools next year will be D tools, and that has a better gross margin of around 50% in comparison to the previous tool that had around 40%. So that will give some tailwind on that front. In the EUV margin, I mentioned that we leave that around '24, '25 timeframe. The EV gross margin will be at the corporate gross margin level. Last year it was 0. This year, we expect a significant improvement on that front. I think I reported that in the first 2 quarters, we're looking at about 25% gross margin. But of course, that will continue to grow to that 50 -- to that 50-ish number. So that will give you some improvement. Then, of course, a bit of a question will be the mix of the DPV tools for next year. This year, very, very strong on immersion. Hope to sustain that next year, but it could be a little bit different. But nonetheless, I think the main drivers are such that the gross margin improvement for next year should be possible.

Roger, another question, if I may, on capital allocation very quickly. You've announced a new buyback program. Why wouldn't you instead use the money to buy strategic assets even in your own supply chain or pursue adjacent opportunities, given your market share is well beyond 90%, almost 95% in lithography? Why still emphasize capital return?

Yes, Amit, to be honest, it's not the right question. Maybe your question is obviously the right question. But the right strategic question is not -- you have the money, what are you going to do with it. That's not the way I would approach it. So the way I would approach it is to say where do we want to put strategic emphasis in the next couple of years. And then if we put that emphasis, then I'm pretty sure that we'll be able to finance it, even -- either through free cash flow or other ways that we would be able to do it. So I think the exam question is do we, from a strategic perspective, want to do some of the things that you just mentioned, i.e., take a better share in the -- a bigger share in the supply chain or more move towards adjacencies. And we look at this very strategically. As it relates to the supply chain, we have a very strict policy to say what is it that we want to control ourselves, that we want to have control over as ASML based on how specific the technology is, how differentiating this technology is, et cetera, and what we don't want to do that. And where we're not compelled because of the added value, because of the uniqueness, because of the complexity of the technology, where we are not compelled that needs to be in-house, we really try to keep it outside of the house and in other ways make sure that our interests are best served by the supply chain. And I think we're pretty happy with the way we have it today. And because of that, we don't plan for significant acquisitions in the supply chain at this stage. Of course, once in a while, something might pop up, but we're not planning for very significant acquisitions there. On the adjacencies, I think we've said it before, our focus right now really is making sure that the demand that is out there in the market on the scanner side and on the metrology and inspection side is being met. We still have significant further industrialization to be done and bringing high-NA up and running. And we think those 2 objectives really require the undivided attention of management. And if we were going to enter into new markets for us, we think that would -- that attention could be diluted, and we believe that is not in the interest of the value of the company and the interest of the shareholders. So that's why at this stage, we're not envisaging, at least not in the medium term, very significant investments to be made outside of the CapEx and the R&D that we would do for the focus areas that I just talked about. And that means that in all likelihood, very significant free cash flow will be available for distribution to shareholders. So to me, the exam question is slightly different. I want to start with what are the strategic objectives and then simply conclude that there is quite some free cash flow available and that will be distributed. I mean, beyond '25, once high-NA is up and running, alternative routes will definitely be looked into. But all the way through '25, our undivided attention is with high-NA and further industrialization of [ loaning ].

Got it, Roger. I must confess, I should have been mindful in addition to being a CFO, you're also a professor. So have to be correct with my questions. Well, in the interest of time...

[indiscernible]

In the interest of time, I was hoping to probably I'll round off with just one big picture question instead of 2, which is talent management is, of course, the lifeblood of a successful organization, particularly one as innovative as yours. There's clearly a war for talent out there. What are you doing to attract talent? Will we see a change in the way ASML incentivizes its workforce, greater stock options? How do you go about attracting talent, the best and the brightest to Veldhoven?

Yes. Amit, it's a great question, just to compensate for my comment on the previous question. This is a great question. And frankly, I mean, it's something that requires a lot of our attention. Here in Veldhoven, we've seen a lot of growth, obviously, in the organization, a lot of growth in the talent that we employ. But I think it's also fair to say that with the position that we have within the Netherlands and within the Veldhoven community, our ability to attract talent is actually at a pretty high level, right? So we work very, very closely with all the technical universities, not just in the Netherlands, but also in the region and far beyond that. And our ability to attract talent, I think, is at a pretty good level. So that's why we've been able to attract a big chunk of people, but also at the levels of -- with the competencies that we were looking for. Outside of Veldhoven, of course, it's -- we compete with many others. When we're on the West Coast, when we're on the East Coast in the U.S., when we're in Taiwan, in Korea, of course, there are many, many players in our ecosystem that we're competing against in terms of the labor market. I think our value proposition is still great. I mean the challenges that we can give our people are phenomenal. The growth potential for people within our organization, for an organization that grows still rapidly is battling with such big, intrinsic questions. I think that is very appealing to people. Of course, on the remuneration side, we've just got to make sure that we're competitive there as well. So far, I think we've been able to manage that without getting to dramatic shifts in some of the things that you just mentioned in terms of ownership options or whatever. So, so far, we find that our proposition is strong enough to compete in this market at conditions that are not dramatically different from the conditions that we've been able to offer in the past couple of years.

Got it. With that, we are out of time. On behalf of Citi and its clients, thank you for joining us, Roger, and best wishes to you and the rest of the ASML team. Ladies and gentlemen, this concludes the keynote session on ASML.